An Account Concerning the Resolution of Equations in Numbers ; un parted by Mr. Iohn Collins .
This Account should have been annexed to what was discoursed of Monsieur Slusius his Mesolabe in the precedent Tract , if then we had found room for it .
For , the Reader having there understood , how far the Geometrick part of Algebra is advanced by that excellent person , it was likely , he would be inquisitive to hear somewhat concerning the Exegesis Numerosa , or the Resolution of AEquations in Numbers .
For whose satisfaction herein , we shall here insert the Account then omitted , being part of a narrative , formerly made by M. Iohn Collins touching some late Improvements of Algebra in England , upon the occasion of its being alleged , that none at all were made since Des Cartes .
1 .
It hath been observed by divers of this Nation , that in any AEquation , howsoever affected , if you give a Root , and find the Absolute number or Resolvend ( which Uieta calls Homogeneum Comparationis ) and again give more Roots and find more Resolvends , that if these Roots or rather rank of Roots be assumed in Arithmetical progression , the Resolvends , as to their first , second or third differences , &amp;c. imitate the Laws of the pure Powers of an Arithmetical progression of the same degree , that the higest Power or first term of the Equation is of .
e. g. In this Equation aaa-3 aa+4 a=N , 1.dif 2.dif 3.dif .
10 Then N. or the 740 218 If a be= 9 Absolutes or Re522 48 8 170 6 solvends will be 352 42 128 6 7 224 36 found to be 92 6 132 To wit the 3d .
differences of those Absolutes are equal , as , in the Cubes of an Arithmetical Progression .
2 .
To find , what habitude those differences have to theCoefficient , of the Equation , 'ist best to begin from an Unit .
3 .
In any Arithmetical Progression , if you multiply NumP ppp bers bers by pairs , you shall create a rank of Numbers whose 2d .
differences are equal ; and if by ternaries , then the 3d .
differences of those Products shall be equal .
And how to find the greatest Product of an Arithmetical Progression of any number of terms having any common difference assigned , contained in any Number proposed , is showed by Pascal in his Tract du Triangle Arithmetique , where he apply 's it to the Extraction of the Roots of simple powers .
4 .
It appears , How this rank may be carried easily by Addition , till you have a Resolvend either equal or greater or less , than that proposed .
5 .
When you have a Majus and Minus , you may interpole as many more terms in the Arithmetical Progression as you will , that is to say , Subdivide the Common difference in the Arithmetical Progression , and render it less ; and then renew , and find the Resolvends , which are easily obtained out of the Powers and their Coefficients , which are supposed known , and may be readily raised from a Table of Squares and Cubes , &amp;c. with which kind the Reader may be furnished in Guldini Centrobaryoa and Babingtons Fireworks : By this means you may obtain divers Figures of the Root ; and then the General Method of Vieta and Harriot runs away more easily , and is so far improved , that after any figure is placed in the Root , most certain Characters are given to know by aide of the subsequent Dividend and Divisor , Whether the figure before assumed be too great or too small : or lastly it may well be concluded , that , as in Logarithmes , when you propose such an one as is not absolutely given in the Canon , you doe by Proportional Work , using the aid of their first differences ( when their Absolute Numbers differ by Unit ) find the absolute Number true to 5 .
or 6 .
places further than the Canon gives it(the reason whereof is , that the first Differences doe likewise agree to about the same Number of places ; ) that I say , the like may be done in AEquations , after divers of the first figures of the root are found ; provided there be the like agreement in the first differences of the interpoled Resolvends .
Moreover we ought here to take notice of a more Subtle kind of Interpolation , common to all gradual Ranks or Progressions of Numbers , wherein Differences happen to be equal : Of which kind the Reader may find Examples in Briggit Arithmetica Logarithmica et Trigonometria Britannica , relating to Logarithmes , Sines , and the Powers of an Arithmetical Progression : But the method there delivered may be rendred more easy and general , viz. by aid of a Table of figurat Numbers , by deriving Generating differences sought , from those given ; a doctrine , that easily flows from Mercators Logarithmotechnia , and of use in the Case in hand , should we suppose these Powers and their Coefficients unknown , or a Table of Squares and Cubes wanting , and give nothing more , than a few Resolvends belonging to equal Moments or Spaces .
And this may likewise be of good use in Guaging , when having the Contents of a Solid , for every 3 .
Inches , more or less , given , without knowing the dimensions of the Figure , and even in most Cases , when the differences are Progressive of one kind , without knowing the Figure it self , having nothing given but its Contents at several equal Parallel distances , each such distance may be subdivided , and made as many as you please , and the respective Contents found by this general Method of Interpolation .
* After one Root is obtained , the Methods of Huddenius and others will depress the Equations so as to obtain more , and consequently all of them .
6 .
It is easy , by a Table of figurate Numbers to give the sum of any such Rank or any term in it relating to a known part of the Series of Equals or Roots ; but e converso , giving the Resolvend to find the Root , coms to an Equation as difficult as that proposed ; as in D. Wallis his Chapter of Figurate Numbers .
* Nota .
The Author ( M. Collins ) having explained Mercators Logarithmotechnia in English , and illustrated the elegant Doctrine thereof with Examples , hath likewise handled this work of Interpolation , and makes the Logarithmes true to 25 .
or 30 .
places of figures by mere Division ( or Proportion ; ) having herein advanced that Author 's doctrine thereof by Division , which ( as it is there illustrated ) did not seem to extend far enough .
This hath already been communicated , some Months since , to some of the Members of the R. Society , and may be expected hereafter .
7 .
Some affirm , they can give good Approaches for the obtaining a Root of any purepower , affected Equation , or for the finding of any of the mean Proportionals in any Rank between two extremes given .
8 .
Others pretend to have found out a method ( incited thereto by an example in Albert Gerards Invention Nouuelle en Algebre a Amsterdam 1629 .
) so much , by comparing of Equations , to inP ppp2 crease crease or diminish the unknown Root of Equation , as to render it a whole number ( or less differing therefrom , than any Error assigned , ) and by Albert Gerards Method of Aliquot parts to find the same , and thereby the Root sought , although it be a Mixed Number , Fraction , or Surd .
Probably this may sympathise with what is promised by the Learned Huddenius in Annexis Geometriae Cartesianae , where he saith , he intended not then to publish certain Rules , he had ready ; whereof one was , To find out all the irrational Roots both of Literal and Numeral Equations .
This must be understood when such Roots are possible ; for it is certain , there are infinite Equations , whose Roots are no ways explicable , either in whole or mixed numbers , Fractions or Surds , and can be no otherwise explained , but by a quam proxime .
9 .
The Author of this Narrative considering , that the Conick Sections may be projected from lesser Circles placed on the Sphere , and thence easily ( otherwise than hitherto hath been handled ) described by Points , and that by their Intersections some Spherick Problem is determined , accordingly he found , that this following Problem according to the various Scituation of the Eye , and of the Projecting Plain , would take in all Cases .
The Distances of an unknown Star are given from two Stars of known Declination and Right Ascension ; the Declination and Right Ascension of the unknown Star is required .
And saith , he hath observed , that , admitting the Mechanisme of dividing the Periphery of a Circle into any number of equal parts , or ( which is equivalent ) the Use of a Line of Chords , that this Problem , wherever the Eye be placed , may be resolved by Plain Geometry , and yet the Ey shall be so placed , as to determine it by the Intersections of the Conick Sections ; consequently those Points of Intersection ( the Species and Position of the figures being given ) may be found without describing any more Points than those sought ; and the Lengths of Ordinates falling from thence on the Axes of either figure calculated by mixed Trigonometry , and hence likewise the Roots of all Cubick and Bi-quadratick Equations found by Trigonometry .
For giving from the Mesolabe mentioned the Scheme that finds these Roots , it will then be required to fit those Sections into Cones , which have their Vertex either in the Center ; or an assigned point in the Surface of the Sphere , to which they relate as projected , and proceed to the resolution of the Problem proposed : And how to fit in those Sections , see the 7 .
books of Apollonius , Mydorgius , the 3d .
Volume of Des Cartes 's Letters , Leotaudi Geometrica Practica , Andersonii Exercitat .
Geometricae .
As to the Problem it self , it is determined on the Sphere by the Intersections of the two lesser Circles of Distance , whose Poles are the known Stars .
And this Problem hath divers Geometrick ways of resolution .
1 .
By Plain Geometry ( in the sense before mentioned ; ) Supposing a Plain to touch the Sphere at the North pole : if the Eye be at the South-pole , projecting those Circles into the Said Plain , they are still Circles ( by reason of the sub-contrary Sections of the Visu l Cones ) whose Centers fall in the sides of the Right-lin 'd Angle , made by the Projected Meridians , that pass through the known Stars ; and thus the Problem is easily solved in this manner .
2 .
If it be required to be performed by Conick Geometry ; in one case it may be done , by placing the Ey at the Center of the Sphere , and projecting as before ; to wit , when the longer Axes of the figures being produced concur above the Vertex ; Here the Problem is determined by the Intersections of two Conick Sections ( whereof a Circle cannot be one , unless its Center be in the Axis of the other figure .
) And in this second Case these points of Intersection fall in the same right line or projected Meridian , they did before , but at a more remote distance from the Pole-point , to wit , in the former Supposition , the Solar distance was measured by a Right line , that was the double Tangent of half the Arch ; here it is the Tangent of the whole Arch. Hence it is evident , how one Projection may beget another , yea infinite others , altering the Scale ; and how the lesser Circles in the Stereographick Projection help to describe the Conick Sections in the Gnomonick Projection : But ( to reduce the matter to one common radius ) if we suppose two Spheres equal , and so placed about the same Axis , that the Pole-point of the one shall pass through the Center of the other , and the Touch-plain to pass through the said Center or Pole-point , and that a lesser Circle hath the same position in the one as in the other ; Then , if the Ey be at the South-Pole of the one , it is at the Center of the other ; and any projected Meridian drawn from the projected Pole-point to pass through both the projections of these lesser ser Circles , the distances of the Points of intersection are the Tangents of the half and the whole Arch of the Meridian so intersected .
But as to the Points of Intersection , which determine the Problem proposed , they may be found without the aid of the former way , from a Gnoonmick and Stereographick method of measuring and setting off the sides and angles of Spherical Triangles in those Projections , which is necessary in what follows .
3 .
If the Problem is to be performed by Mixed Geometry , as by a Circle and either a Parabola , Hyperbola , or Ellipsis , the Circle may be conceived to be the Sun-contrary Section of a Cone projected by the Eye at the South-pole , and any of the rest of the Sections by the Eye at the Center of the Sphere .
4 .
If by any of the Conick Sections however posited ; the projecting Plain may remain the same , but the Eye must be in some other part of the Surface of the Sphere , and not in the Axis .
These things were mention d to invite the Learned to their Consideration : I shall only further add , that we cannot say , what may be expected from the labours and endeavours of divers Learned men of this Nation , particularly from Dr. Wallis , who hath so excellently resolved and constructed all Cubick AEquations at the end of the first Treatise of his Opera Mathematica by aide of a Cubick Parabolaster , mentioning , that by such Curves the Roots of all AEquations may be found : And who hath promised a Treatise of Algebra and Angular Sections , wherein the Reader need not doubt to meet with satisfaction in these Mysteries .
Nor ought we to omit the mentioning of the Modest and Learned Mr. Barrow , who ( among many other excellent Subjects , and particularly his Opticks now , at the Press ) hath performed , what the famous Italian Geometer Mich. A. Ricci hath promised in Exercitat .
Geometrica ( printed at Rome 1666 .
and lately reprinted here ) about Curves of several degrees , that serve to determine and resolve all AEquations : which hath likewise been done by other Learned men of this Nation .

